Hydrogen peroxide – sensitive enzyme sensor based on phthalocyanine thin film

Hydrogen peroxide – sensitive enzyme sensor based on phthalocyanine thin film

Analytica Chimica Acta 391 (1999) 289±297 Hydrogen peroxide ± sensitive enzyme sensor based on phthalocyanine thin ®lm T.A. Sergeyevaa,*, N.V. Lavrik...

196KB Sizes 0 Downloads 125 Views

Analytica Chimica Acta 391 (1999) 289±297

Hydrogen peroxide ± sensitive enzyme sensor based on phthalocyanine thin ®lm T.A. Sergeyevaa,*, N.V. Lavrikb, A.E. Rachkova, Z.I. Kazantsevab, S.A. Piletskya, A.V. El'skayaa a

Institute of Molecular Biology and Genetics, National Academy of Sciences of Ukraine, 150 Zabolotnogo Str., 252143 Kiev, Ukraine b Institute of Semiconductors Physics, National Academy of Sciences of Ukraine, 45 Prospekt Nauki Str., 252650 Kiev, Ukraine Received 30 September 1998; received in revised form 10 February 1999; accepted 21 February 1999

Abstract An enzyme biosensor speci®c for hydrogen peroxide was developed using a new conductometric transducer based on tetratert-butyl copper phthalocyanine (ttb-CuPc) thin ®lms and horseradish peroxidase as sensitive element. This analytical system is based on detection of molecular iodine produced as a result of the oxidation of the iodide ions by hydrogen peroxide in the presence of horseradish peroxidase. For the detection of the peroxidase-initiated reaction the ability of the ttb-CuPc thin ®lm to change its conductivity in response to the appearance of molecular iodine is used. To minimise the interfering effect of the aqueous electrolyte on the conductometric response of the ttb-CuPc thin ®lm itself, gold interdigitated electrodes bearing ttbCuPc layer were covered with a hydrophobic gas-permeable membrane. Thermally evaporated calixarene or plasma polymerised hexamethyldisiloxane was used as a gas-permeable membrane material. In order to assess the optimum sensor technology as well as the operating regime, impedance spectroscopy data were analysed. For biosensor creation horseradish peroxidase was deposited on the sensitive part of the electrodes in a cross-linked bovine serum albumin matrix. The possibility of hydrogen peroxide detection with the biosensor proposed in the range 5±300 mM was demonstrated. The operational stability of biosensor was at least 7 h and the relative standard deviation did not exceed 10%. When stored at ‡48C the sensor response was stable for more than 90 days. The dependencies of the sensor response on pH, buffer and NaCl concentrations were investigated. # 1999 Elsevier Science B.V. All rights reserved. Keywords: Horseradish peroxidase; Biosensor; Hydrogen peroxide detection; Interdigitated planar electrodes; Tetra-tert-butyl copper phthalocyanine; Langmuir±Blodgett ®lms

1. Introduction Development of reliable methods of hydrogen peroxide determination is of great importance in both biological and industrial ®elds. Detection of hydrogen peroxide is important from several points of view: *Corresponding author. Tel.: +380-044-266-07-49; fax: +380044-266-07-59; e-mail: [email protected]

1. it is widely used in food industry as a sterilising agent and may be present in ®nal products which can lead to the loss of nutritional value and appearance of toxic compounds such as hydroperoxides, epoxides and aldehydes [1]; 2. hydrogen and organic peroxides can be released in the environment from industrial processes and during the ozonation of drinking water [2,3];

0003-2670/99/$ ± see front matter # 1999 Elsevier Science B.V. All rights reserved. PII: S 0 0 0 3 - 2 6 7 0 ( 9 9 ) 0 0 2 0 3 - 2

290

T.A. Sergeyeva et al. / Analytica Chimica Acta 391 (1999) 289±297

3. measurements of hydrogen peroxide produced during the reaction of many important chemicals in the presence of oxidases forms the basis for several biological and medical tests [4±6]; 4. hydrogen peroxide is frequently used in pharmaceutical and cosmetic formulations [7]; 5. horseradish peroxidase is widely used as label in enzyme-linked immunosorbent assays (ELISA). Conventional methods for the determination of hydrogen peroxide including spectrophotometry [8], colorimetry [9] and chemiluminescence [10] often involve complicated methods and expensive equipment. A commonly used electrochemical method for the detection of hydrogen peroxide is its oxidation at high overpotentials at a wide variety of electrodes [11± 17]. The main disadvantage of this approach is its poor speci®city because the high voltage applied to the working electrode can lead to unacceptable interferences. To reduce the interfering effect caused by substances which can be oxidised at used potentials, alternative systems which can operate at much reduced voltages are being developed for hydrogen peroxide detection. A number of amperometric enzyme sensors combining the speci®city of enzymatic reaction with the high sensitivity of electrochemical transducer have been proposed for this purpose [18±22]. An amperometric sensor based on horseradish peroxidase (HRP) modi®ed platinised carbon particles which are capable of reducing hydrogen peroxide at 0.0 V (vs. SCE) has been developed by Cardosi [23]. A new scheme for fabricating bulkmodi®ed amperometric biosensors based on graphite±PTFE electrodes which give a possibility of hydrogen peroxide detection at 0.0 V (vs. SCE) with a detection limit of 2.5 mM, and a linear dynamic range of 2.5±150 mM has been reported [24]. A potentiometric sensor for this purpose has been proposed by Zul®car et al. [25]. The potentiometric response of this sensor in the ¯ow system was linear in the concentration range 0.75±50 mM. As one can conclude considerable progress and practical achievements in the ®eld of hydrogen peroxide determination have been obtained by using amperometric and potentiometric biosensors. Until now no conductometric biosensors for this purpose have been reported. The main reason for this is that because of the relatively low impedance of aqueous

medium signi®cant dif®culty is created in proper operation of the conductometric biosensors. Nevertheless, conductometric biosensors can be more easily integrated since they do not need a reference electrode and the conductometric transducers can be manufactured using simple thin ®lm technology [26±28]. In the present study, the possibility of hydrogen peroxide monitoring was investigated with a peroxidase enzyme sensor based on a conductometric transducer having tetra-tert-butyl Cu-phthalocyanine thin ®lms as a sensitive element. The construction of the transducer is simple, of low cost and has good possibilities for mass production. 2. Experimental 2.1. Materials Tetra-tert-butyl copper phthalocyanine (ttb-CuPc) of high purity was purchased from the Research Institute of Organic Dyes (Moscow, Russia). Chloroform and benzene of a reagent grade were obtained from Sigma and were distilled prior to use. Hexa-tertbutyl-calix[6]arene was synthesised at the Institute of Organic Chemistry (Kiev, Ukraine) in Dr. Kalchenko's laboratory as described elsewhere [29]. Hexamethyldisiloxane was obtained from the microelectronics plant ``Kvazar'' (Ukraine). Glutaraldehyde and bovine serum albumin (BSA) were purchased from Sigma. HRP (EC 1.11.1.7) with an activity 257 U/mg (guaiacol as a reducing substrate) was purchased from Biozyme. Hydrogen peroxide standardised by iodimetric titration was freshly prepared in deionised water. Iodine (I2) was purchased from Sigma as crystals of 99% purity. I2 crystals were dissolved in 96% (v/v) ethanol to prepare a 0.2 M stock iodine solution for further experiments. Potassium iodide (KI) was obtained from Sigma and a 0.5 M stock solution of KI was prepared by dissolving in 0.02 M K-phosphate buffer, pH 6.0 immediately before use. 2.2. Iodine-sensitive transducer and measurements of the I2 concentration Interdigitated microelectrodes were purchased from Emokon (Kiev, Ukraine). The ceramic chips of

T.A. Sergeyeva et al. / Analytica Chimica Acta 391 (1999) 289±297

5 mm30 mm size and 0.2 mm thickness contained two identical pairs of gold interdigitated electrodes. The electrodes were deposited onto a ceramic substrate by using conventional thin-®lm technology which includes the following main stages: 1. vacuum deposition of 0.1 mm chromium adhesive layers onto ceramic substrates; 2. vacuum deposition of 1 mm gold on the top of the chromium layers; 3. patterning of the interdigitated microstructures by photolithography. Each finger in the electrode arrays was 20 mm wide and approximately 1 mm long. The separation between the adjacent fingers was 20 mm. The total sensitive area of each electrode array was about 1.5 mm2. Iodine-sensitive transducers were prepared by deposition of thin ttb-CuPc ®lm onto the active area of the interdigitated electrodes by one of the following methods: casting from chloroform/benzene solutions, Langmuir±Blodgett (LB) deposition, and thermal evaporation in vacuum. The details of the thermal evaporation and LB deposition procedures are given elsewhere [30,31]. Hydrophobic gas-permeable membranes (HGPMs) deposited on the ttb-CuPc ®lms were formed by either thermal evaporation of hexa-tertbutyl-calix[6]arene (HGPM-1) or plasma polymerisation of hexamethyldisiloxane (HGPM-2) as described elsewhere [32]. The cross-section of the resulting structures is schematically shown in Fig. 1. The thickness of each of the deposited organic layers was de®ned by ellipsometric measurements using Simonocrystal satellites.

291

To measure impedance spectra and conductometric responses of the sensor, the differential pair of the interdigitated electrodes with the ttb-CuPc/HGPM layer was connected to the electronic units as described earlier [33]. The output voltage was recorded with a chart recorder and the values corresponding to the steady-state response were taken as a sensor response. These values were also used for impedance analysis. To plot impedance spectra, the absolute values of the complex impedance |Z| of the sensor were calculated using the following equation: jZj ˆ Ro …Ug ÿ Ux †=Ux ;

(1)

where Ro is the resistance across the a.c. nanovoltmeter input, Ug the a.c. generator output voltage (60 mV) and Ux is the voltage measured with the a.c. nanovoltmeter. 2.3. Enzyme immobilisation and measurements of H2O2 concentration Horseradish peroxidase was immobilised onto the transducer surface in a cross-linked BSA matrix according to the following procedure. Peroxidase and BSA were dissolved separately in 0.1 M K-phosphate buffer solutions, pH 8.0. The resulting protein concentrations were 100 mg/ml in both solutions. Equal aliquots of these solutions were mixed with inositol (Sigma, St. Louis, USA) and ZnSO4 so that the concentrations of the inositol and ZnSO4 in the resulting mixture were 5% (w/v) and 0.9 mM, respectively. Approximately 1 ml of this mixture was deposited on a sensitive area of the electrodes. To complete the polymerisation of the membrane, the sensor chip was exposed to glutaraldehyde vapours for 30 min and then dried at room temperature for 15 min. For H2O2 determination, the prepared enzyme sensor was placed in a 2.5 ml glass cell with gently stirred 0.02 M K-phosphate buffer (pH 6.0) containing 0.01 M KI and 0.15 M NaCl. The measurements were carried out at room temperature. The sensor responses were recorded after successive additions of the stock H2O2 solution into the buffer. 3. Results and discussion

Fig. 1. The scheme of a conductometric horseradish peroxidasebased enzyme sensor with a chemoresistive layer separated from the region where the enzymatic reaction occurs.

The conductometric method of H2O2 detection is based on the reaction of iodide ions oxidation by H2O2

292

T.A. Sergeyeva et al. / Analytica Chimica Acta 391 (1999) 289±297

in the presence of HRP: ‡ peroxidase

H2 O2 ‡ 2Iÿ ‡ 2H

!

2H2 O ‡ I2

(2)

Therefore, the amount of hydrogen peroxide used in this reaction can be monitored by measurement of either iodide consumption or free iodine release. Iodide-sensitive potentiometric electrodes have been used previously in order to measure the iodide concentration [34,35], but integration of potentiometric biosensors implies quite sophisticated technology. On the other hand, the reaction (2) can be monitored by measurement of the free iodine released. In particular, we propose to monitor the iodine concentration using a conductometric transducer with a phthalocyanine thin ®lm as a sensitive element. To detect the molecular iodine concentration, the ability of ttb-CuPc ®lm to change its conductivity in response to appearance of I2 is used in the present study. Chemoresistors including phthalocyanins are widely recognised as one of the most attractive basis for chemical and biosensors due to their direct electrical responses and good compatibility with electronics [36]. However, high electrical conductivity of aqueous electrolyte prevents realisation of biosensors on this basis, therefore properties of chemoresistors were mainly studied in gaseous environments only [30,31]. A speci®c feature of the conductometric transducer designed is that ttb-CuPc ®lm was covered by a hydrophobic gas-permeable membrane (HGPM) to suppress interfering effect of an aqueous electrolyte (Fig. 1). To de®ne the optimum technological procedures and operating frequency, impedance spectra of a set of the prepared structures were measured in air (Fig. 2, curve 1) and in various electrolytes, with and without molecular iodine (Fig. 2, curves 2±4). As can be seen, at the lowest frequency, the increase of the electrolyte conductivity has no signi®cant effect on the impedance (Fig. 2, curves 2 and 3), especially in the case of LB-deposited ttb-CuPc ®lms. At the same time, lowfrequency impedance decreases when 20 mM of free iodine is present in the electrolyte (Fig. 2, curve 4). It was shown that the shift of the impedance spectra due to the presence of iodine is much lower in the case of the casted ®lms. Therefore, samples with the latter type of ttb-CuPc ®lms were eliminated from further experiments. The structures with the LB-deposited and thermally evaporated ®lms exhibited quite similar characteristics. Thermal evaporation of the ttb-CuPc

®lms provides all-dry processing of the transducers up to biomaterial immobilisation. On the other hand, the LB deposition uses more effectively the material to be deposited. To minimise the interfering effect of the electrolyte, the working frequency of 1 Hz was selected for monitoring the sensor impedance in all subsequent experiments. The chemoresistors with a LB-deposited and thermally evaporated ttb-CuPc layer demonstrated similar sensitivity to the appearance of free iodine in solution. Similar results were obtained in the case of the thermally evaporated active layers. The response was rather slow and at least 10 min was needed to reach the steady-state response. As noted before [29,30] such a slow kinetics is typical for conductivity changes in phthalocyanine ®lms at room temperature. Similar responses on iodine were obtained for both types of hydrophobic gas permeable membranes, HGPM-1 and HGPM-2, of equal thickness (100 nm). However, HGPM-2 was chosen for the further experiments because of its better adhesion and mechanical stability. HRP was immobilised on a surface of the designed iodine-sensitive transducer in a cross-linked BSA matrix. To compensate non-speci®c reactions of the biosensor, an enzyme-free BSA matrix was deposited on a reference transducer. All measurements were carried out in a differential pair mode. The calibration curve of the biosensor obtained in 20 mM potassiumphosphate buffer, pH 6.0 was found to be linear within the range of the hydrogen peroxide concentration from 5 to 300 mM (Fig. 3), each point represents the average of three measurements and the sensor standard deviation did not exceed 10%. However, no sensor response was observed when H2O2 was added to the buffer solution in the absence of KI. Addition of the iodide in the absence of the substrate, H2O2, did not cause any noticeable response as well. The dependence of the enzyme sensor response on the pH of the sample solution was investigated. The greatest response takes place at a pH between 5.0 and 6.5 (Fig. 4). This corresponds approximately to the region where the peroxidase activity has a maximum. One can therefore conclude that the observed responses are indeed controlled by the rate of the enzymatic reaction (2). The generally recognised main drawback of conductometric biosensors is their poor speci®city exhib-

T.A. Sergeyeva et al. / Analytica Chimica Acta 391 (1999) 289±297

293

Fig. 2. Impedance spectra of iodine-sensitive conductometric transducer with casted (a) and Langmuir±Blodgett deposited (b) tetra-tert-butyl copper phthalocyanine films. The impedance spectra were measured for: 1 ± dry electrodes, 2 ± electrodes immersed in distilled water, 3 ± 0.15 M NaCl solution, 4 ± 0.15 M NaCl solution containing 20 mM I2.

ited particularly in measurements in buffers with a high capacity and ionic strength. The dependencies of the buffer capacity and ionic strength on the magnitude of the peroxidase enzyme sensor response have

been investigated. It has been shown that neither ionic strength nor buffer capacity affected the magnitude of the sensor responses signi®cantly (Figs. 5 and 6). This fact seems to be of great importance because it gives a

294

T.A. Sergeyeva et al. / Analytica Chimica Acta 391 (1999) 289±297

Fig. 3. Dependence of horseradish peroxidase-based enzyme sensor response on substrate concentration. Measurements were carried out in 0.02 M potassium phosphate buffer, pH 6.0, containing 0.15 M NaCl and 0.01 M KI.

Fig. 4. The pH dependence of the activity of horseradish peroxidase immobilised on the surface of the iodine-sensitive transducer. Measurements were carried out in 0.02 M potassium phosphate buffer containing 0.15 M NaCl, 50 mM H2O2 and 0.01 M KI.

T.A. Sergeyeva et al. / Analytica Chimica Acta 391 (1999) 289±297

295

Fig. 5. Dependence of horseradish peroxidase-based enzyme sensor response on NaCl concentration in working solution. Measurements were carried out in 0.02 M potassium phosphate buffer containing 50 mM H2O2 and 0.01 M KI.

Fig. 6. Dependence of horseradish peroxidase-based enzyme sensor response on buffer concentration of working solution. Measurements were carried out in potassium phosphate buffer containing 0.15 M NaCl, 50 mM H2O2 and 0.01 M KI.

296

T.A. Sergeyeva et al. / Analytica Chimica Acta 391 (1999) 289±297

Table 1 Characteristics of HRP-based biosensor Operation mode Optimum pH Linear dynamic range Baseline recovery time Relative standard deviation Storage stability Operational stability

4. Conclusions Steady state 6.0 5±300 mM 15 min 10% >90 days 7h

possibility to avoid the main drawback of the conductometric detection method. The results obtained prove that the proposed transducers have high and speci®c sensitivity to the iodine generated by the enzymatic reaction. Thus, despite some slight effect of the electrolyte on the measured sensor impedance, the hydrophobic membrane is thick enough to provide stable operating of the sensor in the range of physiological pH and ionic strength. Stability of a biosensor system is a very important characteristic for their commercial application. The operational stability test demonstrated that the steadystate response was not decreased for at least 7 h (approximately 30 measurements). When stored at ‡48C, the biosensor response was stable for at least three months. The main characteristics of the horseradish peroxidase-based biosensor are summarised in Table 1. Both the response and recovery times could be signi®cantly reduced if the initial rate of the conductivity increases was measured instead of steady state responses. However, this operation mode would also lead to an increased error due to a much lower signal/ noise ratio in the beginning of the response kinetics. The range of the detectable concentrations can be extended to the upper end by adding a diluting system based, for instance, on a micro¯uidic device. At the same time, the threshold sensitivity of the proposed sensor has to be still improved to make it useful for biodiagnostic analysis of hydrogen peroxide. One of the possible ways to achieve this is optimisation of diffusion properties of the matrix for the enzyme immobilisation. The stability of a biosensor over a long period of time is one of the most critical characteristics that de®nes the potential for commercial applications. Although 90 day shelf-lifetime requires refrigerating at ‡48C, this can be considered as appropriate for most of the practical applications.

A chemoresistor based on the organic semiconductor material, phthalocyanine, has been used in aqueous medium as a transducer for biosensors. Both the thermal evaporation and Langmuir±Blodgett deposition of the chemoresistive material provided transducers with satisfactory and quite reproducible parameters. Due to the presence of hydrophobic gas-permeable membrane covering the iodine-sensitive ®lm, neither ionic strength nor buffer capacity has a signi®cant in¯uence on the measured responses. The separation of the chemoresistive material and the aqueous medium effectively reduces electrolyte interference, and thus, practically eliminates the main disadvantage of the conductometric method as applied to biosensors. It has been shown that free iodine generated due to the peroxidase-initiated reaction can be effectively monitored with the designed transducers. Acknowledgements Financial support from Ministry Ukraine for Science and Technology (Grant 05.41.07/005-92) is gratefully acknowledged. The authors thank Tammy Calvert for linguistic advice. References [1] M.G. Simic, M. Karel (Eds.), Autoxidation in Foods and Biological systems, Plenum Press, New York, 1979. [2] IARC Monographs on the evaluation of the carcinogenic risk of chemicals to humans. Allyl compounds, aldehydes, epoxides and peroxides, IARC, Lyon, France 36 (1985) 267. [3] W.H. Glaze, Environ. Sci. Technol. 21 (1987) 224. [4] L. Braco, J.A. Daros, M. de la Guardia, Anal. Chem. 64 (1992) 129. [5] G.F. Hall, A.P.F. Turner, Anal. Lett. 24 (1991) 1375. [6] R.Z. Kazandjian, J.S. Dordick, A.M. Klibanov, Biotechnol. Bioeng. 28 (1986) 417. [7] J. Wang, Y. Lin, L. Chen, Analyst 118 (1993) 277. [8] R.M. Sellers, Analyst 105 (1980) 950. [9] Y. Ito, Y. Tonogai, H. Suzuki, J. Assoc. Off. Anal. Chem. 64 (1981) 1448. [10] G.M. Kok, T.P. Holler, M.B. Lopez, Environ. Sci. Technol. 12 (1978) 1073. [11] R.M. Iannielo, A.M. Yacynych, Anal. Chem. 53 (1981) 2090.

T.A. Sergeyeva et al. / Analytica Chimica Acta 391 (1999) 289±297 [12] J. Wang, N. Naser, L. Angnes, W. Hui, L. Chen, Anal. Chem. 64 (1992) 1285. [13] L.C. Clarc, Meth. Enzymol. 56 (1979) 448. [14] H.P. Benetto, D.R. Keyzer, G.M. Delaney, Analyst 8 (1987) 22. [15] X.H. Cai, B. Ogorevc, G. Tavcar, J. Wang, Analyst 120 (1995) 2579. [16] M. Somasundrum, K. Kirtikara, M. Tanticharoen, Anal. Chim. Acta 319 (1996) 59. [17] R. Toniolo, N. Comisso, G. Bontempelli, G. Schiavon, Electroanalysis 8 (1996) 151. [18] M. Corsegove, G.J. Moddy, J.D.R. Thomas, Analyst 113 (1988) 118. [19] P.D. Sanchez, P.T. Blanco, J.M.F. Alfarez, Electroanalysis 3 (1991) 281. [20] G.J. Peterson, Electroanalysis 3 (1991) 741. [21] M.S. Lin, S.Y. Tham, G.A. Rechnitz, Electroanalysis 2 (1990) 511. [22] L. Charpentier, N.E. Murr, Analysis 23 (1995) 265. [23] M.F. Cardosi, Electroanalysis 6 (1994) 89. [24] J. Wang, A.J. Reviejo, L. Angnes, Electroanalysis 5 (1993) 575. [25] D. Zulficar, B. Hibbert, P.W. Alexander, Electroanalysis 7 (1995) 722.

297

[26] S.V. Dzyadevich, V.N. Arkhipova, A.P. Soldatkin, A.V. El'skaya, A.A. Shul'ga, Anal. Chim. Acta 374 (1998) 11. [27] A.A. Shulga, S.V. Dzyadevich, A.P. Soldatkin, S.V. Patskovsky, V.I. Strikha, A.V. El'skaya, Biosensors and Bioelectronics 9 (1994) 217. [28] M.S. De Silva, Yu. Zhang, P.J. Hesketh, G.J. Maclay, S.M. Gendel, J.R. Steller, Biosensors and Bioelectronics 10 (1995) 669. [29] J. Vicens, V. Bohmer (Eds.), Calixarenes, a Versatile Class of Macrocyclic Compounds, Kluwer Academic Publishers, Dordrecht, 1991. [30] C.C. Lesnoff, Phthalocyanins: Properties and Applications, VCH, Weinheim, 1989. [31] A.V. Nabok, Z.I. Kazantseva, N.V. Lavrik, B.A. Nesterenko, Int. J. Electronics 78 (1995) 129. [32] G. Kampfrath, R. Hintche, Anal. Lett. 22 (1989) 2423. [33] T.A. Sergeyeva, N.V. Lavrik, S.A. Piletsky, A.E. Rachkov, A.V. El'skaya, Sensors and Actuators B 34 (1996) 283. [34] J.L. Boitieux, G. Desmet, D. Thomas, Clin. Chem. 25 (1979) 318. [35] R.L. Lenado, G.A. Rechnitz, Anal. Chem. 45 (1973) 826. [36] T.M. Swager, M.G. Marcella, Adv. Mater. 6 (1994) 595.